KR101451174B1 - Mac address learning in a distributed bridge - Google Patents

Abstract

Comprising configuring a network node (30) with at least first and second line cards (32) each having a port (36) to operate as a distributed medium access control (MAC) bridge in a layer 2 network A method is disclosed. Each line card includes a respective forwarding database (FDB, 58). After receiving a data packet from a MAC source address to a port of a network node, the data packet is forwarded to at least the first line card for transmission to the MAC destination address. The MAC source address of the data packet is checked against the record in the FDB of the first line card. If the FDB does not have a MAC source address and an associated record of the port where the data packet was made, then the record is added to the FDB of the first line card and a message informing the second line card of the association to at least the second line card send.

Layer 2 network, distributed medium access control (MAC) bridge, network node, port, MAC address, record, forwarding database (FDB), data packet.

Description

Learning MAC Address in Distributed Bridge {MAC ADDRESS LEARNING IN A DISTRIBUTED BRIDGE}

FIELD OF THE INVENTION The present invention relates generally to communication networks, and more particularly to virtual private LAN services (VPLS) and bridging methods and systems within other distributed bridging systems.

A local area network (LAN) connects both computing systems to a layer two level. The term "layer 2" refers to a second layer of a protocol stack formed by a well-known Open System Interface (OSI), also known as a logical link, a data link, or a medium access control (MAC) layer. Each computing system connects to the LAN through a MAC device. A plurality of LANs may be interconnected using a MAC bridge, which is described in " ANSI / IEEE Standard 802. ID (2004) ", entitled " IEEE Standard for Information Technology, Telecommunications and Information Exchange between Systems, Local and Metropolitan Area Networks, Common Specifications, Part 3: Media Access Control (MAC) Bridges "(the 802.1D standard as well as other IEEE standards mentioned here are available at 'standards.ieee.org/catalog/' ). The MAC bridge according to the 802.1D standard allows MAC devices connected to physically separate LANs to be seen by each other as if they were connected to a single LAN. Such bridges include two or more MAC devices interconnecting bridge ports to separate LANs.

The MAC bridge maintains a forwarding database (FDB) to map the destination MAC address of the packets it receives to the bridge port. This bridge builds a forwarding database by means of a learning process that associates the source MAC address of each incoming packet with the port from which the packet was received. When a bridge receives an incoming packet whose destination address does not exist in the database, the bridge floods (i.e. broadcasts) the packet through all available ports except the port where the packet arrived. Other MAC bridges that do not recognize the destination address will also flood that packet to all relevant ports. Through the flooding mechanism, the packet will eventually move all interconnected bridges at least once, and will ultimately reach its destination.

Recently, various means for transmitting Layer 2 packets such as Ethernet frames have been proposed and developed, through a high-speed, high-performance Layer 3 packet network. Methods for this purpose are described, for example, in "Martini et al.," Encapsulation Methods for Transport Frames Over IP / MPLS Networks (IETF draft-ietf-pwe3-ethernet-encap- . This design, as well as other Internet designs mentioned herein, are available at the Internet Engineering Task Force (IETF), www.ietf.org/internet-drafts. This design defines a mechanism for encapsulating Ethernet traffic for communication over an Internet Protocol (IP) network using other tunneling methods, such as Multi-Protocol Label Switching (MPLS) or Generic Routing Encapsulation (GRE) do.

In accordance with the model proposed by Martini et al., The original Ethernet LAN is connected to the IP network by a provider edge (PE) device, which is interconnected by a tunnel through the IP network. As a result of the encapsulation of the Ethernet frame and associated processing functions, the IP network emulates Ethernet trunking and switching operations and can therefore be handled as an Ethernet "virtual-wire" (PW). That is, from the perspective of a native Ethernet LAN connected to a tunnel over an IP network, each PW is a virtual Ethernet point-to-point connection and emulates a physical connection between the two Ethernet ports. The encapsulation method of 'Martini' can also be used in conjunction with a virtual LAN (VLAN), as defined in the IEEE standard 802.1Q.

In addition to these features, we have described how a large number of people create Virtual Private LAN Services (VPLS) that connect different LANs together over an IP network. Such a method is disclosed in, for example, "Virtual Private LAN Service (IETF draft-ietf-12vpn-vpls-bgp-06.txt, December, 2005)" of "Kompella et al. Private LAN Services over MPLS (IETF draft-ietf-12vpn-vpls-ldp-08.txt, November, 2005).

VPLS (also known as Transparent LAN Service - TLS) provides bridged functionality between multiple sites on a large network. The user is connected to the VPLS through a common Ethernet interface. The PW between the nodes to which the user is connected forms a VPLS entity by itself. All nodes in the VPLS act as virtual bridges. The virtual bridge node has a "virtual port", which is the endpoint of the PW that is part of the VPLS. The interface to which the user is actually connected is the physical port at the network edge. The virtual and physical interfaces are treated the same in terms of frame forwarding and address learning. A single supplier node may participate in a stance, each of which is a plurality of VPLSs belonging to different users. From an end-user perspective, the VPLS network is transparent. The user is mistaken as if the provider network is a single LAN domain. Therefore, user nodes on different physical LANs may be connected together via a VPLS connection to form a layer 2 virtual private network (VPN), which appears to the user as the same Ethernet LAN.

A link aggregation (LAG) is a technique by which a group of parallel physical links between two end points in a data network can be joined together into one logical link (hereinafter referred to as "LAG group"). Traffic transmitted between endpoints is distributed between physical links in a transparent manner to clients that receive and transmit the traffic. For Ethernet networks, the link aggregation is defined in Section 43 of IEEE Standard 802.3, "Access Method and Physical Layer Specifications (CSMA / CD)," 2002, Carrier Sense. Section 43 defines a link aggregation protocol sub-layer that interfaces between the MAC client transmitting and receiving traffic over the aggregated link and the MAC layer of the physical link within the aggregation group. The link aggregation sub-layer includes a distributor function for distributing data frames submitted by the MAC client between physical links in the group, and a selector function for receiving the frame over the aggregated link and for forwarding it to the appropriate MAC client .

Embodiments of the present invention provide rewritten MAC learning methods and network nodes implementing such methods. This method is particularly useful in a node environment configured to serve as a virtual bridge, as well as within a Layer 2 virtual private network, as well as other types of distributed bridge nodes, particularly when multiple ports of a node are combined into a LAG group. However, the principles of the present invention can be applied with modifications, if necessary, to facilitate MAC learning in any distributed MAC learning environment.

In some embodiments of the invention, the network node comprises a plurality of line cards with respective ports and is configured to operate as a virtual MAC bridge in a Layer 2 virtual private network (VPN). (One example of such a VPN is VPLS, as described above.) Each line card is typically served by ingress and egress for data packets and shared by these ingress and egress functions. And each MAC forwarding database (FDB). When the ingress line card receives an incoming data packet over a VPN to one port, it consults the FDB to select the port and line card to which the packet should be forwarded based on the MAC destination address When it is not in the FDB, floods the packet over the port in the VPN).

The gateway line card (or line cards) to which the packet is to be transmitted checks for the MAC source address of the data packet for the record in the FDB that it owns. If the FDB of the sending line card does not contain a record that associates the port of the incoming line card with which the data packet was received with the MAC source address, the sending line card adds the record to its FDB. At the appropriate time, the line card transmits a synchronization message to the remaining line cards informing the association of the MAC source address and the ingress port. Typically, all line cards transmit synchronization messages at any given time, but in some circumstances the synchronization messages can be sent immediately upon entry of a new association in the FDB. After receiving the synchronization message, the other line cards update their MAC FDB, if appropriate.

When the forwarding destination of a packet is a Link Aggregation Group (LAG), MAC member selection (i.e., selection of the link to which the packet is to be forwarded) is typically performed on an Ingress line card. Without the synchronization method described above, other members in the LAG will not receive these packets for transmission, and the FDB of the corresponding line card will not be updated. When such a line card receives an incoming packet, the FDB is incomplete, so the result can always be flooded. The synchronization method described herein overcomes this problem by updating the FDB in all line cards within the LAG group within the node (or for stances that are full VPNs). Typically, when a transmitting line card transmits a data packet over a port belonging to the LAG group, the synchronization message sent by the line card identifies the stance and incoming port that is the VPN. Another line card in the same LAG group (and all other line cards that serve as stanzas that are VPNs) is also used to learn MAC address associations even when these other line cards have not yet received the problem packets from the MAC address. Information is available.

Therefore, according to an embodiment of the present invention,

Configuring a network node having at least first and second line cards with respective ports to operate as a distributed medium access control (MAC) bridge in a Layer 2 network;

Providing a respective forwarding database (FDB) to each of the line cards for holding a record associating a MAC address with a respective port of the network node;

Receiving a data packet specifying a MAC destination address on the network from one of a port of a network node from a MAC source address;

Forwarding a data packet received at a network node to at least a first line card, for transmission to a MAC destination address;

Checking the MAC source address of the data packet for the record in the FDB of the first line card; And

Adding the record to the FDB of the first line card if the FDB of the first line card does not include a record of the MAC source address and the association of the one port on which the data packet was received, To the second line card at least to the second line card.

In one embodiment, the step of transmitting the message comprises periodically transmitting a message at predetermined intervals of time to notify at least the second line card of a new association between the MAC address and each port.

Typically, the method includes receiving a message on a second line card, and, in response to the message, if the record is not already present in the FDB of the second line card, FDB. ≪ / RTI > In the disclosed embodiment, the method further comprises: adding a first type of record added in response to the data packet transmitted on one port of the line card and a second type of second type added in response to the message received on the other port of the line card And displaying the records in each FDB of each line card to distinguish the records. The method also includes associating each aging time with each record, refreshing a record in the FDB in response to an additional packet sent by the line card; And if the record is not refreshed within each aging time, removing the record from each FDB.

In some embodiments, transmitting the message comprises transmitting a synchronization packet from a first line card to at least a second line card through a switching core of the network node. In one embodiment, the step of transmitting the synchronization packet further comprises the step of altering the record in the FDB of the first line card if the record in the FDB associates the MAC source address with a port different from the one port from which the data packet was received And transmitting a synchronization update packet to at least the second line card to indicate that the record has been changed.

In the disclosed embodiment, the first and second line cards have respective first and second ports coupled to a link aggregation (LAG), wherein the step of conveying the received data packet comprises: Wherein the step of transmitting the message comprises identifying the LAG group in the message to inform all line cards that are members of the LAG group of the association. Typically, transmitting the data packet comprises flooding a data packet through a port of a line card when a MAC destination address is not present in the FDB, the data packet comprising only a single of the ports in the LAG group It only floods through the port.

In some embodiments, the network node is configured to operate as a plurality of virtual MAC bridges in a Layer 2 virtual private network (VPN), each virtual MAC bridge being configured to serve a stance that is a respective VPN, The record associating the port with the MAC address is maintained independently for each stanza that is the VPN. In the disclosed embodiment, a stance that is a VPN is a stance that is a VPLS of one of a plurality of VPLSs served by a network node, and the step of sending the message is a VPLS stance in a message to notify all line cards serving a stance that is a VPLS. Lt; / RTI >

Typically, the method further comprises transmitting the additional data packet received from the additional MAC source address to the second line card for transmission over the network, checking the additional MAC source address for the record in the FDB of the second line card Adding an additional record relating to the MAC source address to the FDB of the second line card in response to the additional data packet and sending an additional message to notify at least the first line card of the additional record .

Further, according to an embodiment of the present invention, as a node for network communication,

Switching core; And

Wherein the node comprises a plurality of line cards configured to forward packets through a switching node for operating as a virtual medium access control (MAC) bridge in a Layer 2 network,

A plurality of line cards including at least first and second line cards, each line card including a respective port, each forwarding database for holding a record associating a MAC address with each port of the line card, (FDB)

The line card is configured to receive a data packet specifying a MAC destination address from a MAC source address to one port of one line card and then transmit the at least one first line card , The first line card checks the MAC source address of the data packet for the record in the FDB of the first line card, and the MAC database of the first line card transmits the MAC packet to the one port from which the data packet was received A network communication node, which is arranged to add the record to the FDB of the first line card and to send a message informing the at least second line card to at least the second line card if it does not include an association record of the MAC source address Is provided.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will be more fully understood from the detailed description of the embodiments, taken in conjunction with the accompanying drawings, in which:

1 is a block diagram schematically illustrating a communication system, in accordance with one embodiment of the present invention;

2 is a block diagram schematically illustrating details of a line card in a network node, according to one embodiment of the present invention; And

3 is a flow chart schematically illustrating a MAC learning method according to an embodiment of the present invention.

Figure 1 is a block diagram that schematically illustrates a communication system 20, in accordance with one embodiment of the present invention. A Layer 2 VPN, in the form of a VPLS, is provided to the system 20 to connect MAC user terminals of different parts of the network, including exemplary terminals 22 and 24. In the scenario shown, the terminal 22 is connected to a LAN, such as an Ethernet LAN, and the terminal 24 is connected to a WAN 28, such as the Internet or another layer 3 network. However, the VPLS allows users of terminals 22 and 24 to communicate with each other as if they were connected to the same LAN domain.

Although only two user terminals are shown for simplicity in FIG. 1, a given VPLS may connect a very large number of users, typically at very different locations. Further, although the embodiment is described below only with respect to a single VPLS stance, a plurality of stanches, which are various VPLSs, may be provided to the system 20 to serve various groups of users and organs. The specific configuration of the LAN 26 and the WAN 28 is illustrated in FIG. 1 for purposes of illustration and the principles of the present invention may be applied to any network configuration that supports the definition of a Layer 2 virtual private network.

In the exemplary configuration shown in FIG. 1, the network node 30 connects the LAN 26 and the WAN 28. The node 30 includes a plurality of line cards 32 connected by a switching core 34. The line card 32 has a port 36 that connects to other nodes in the LAN 26 and the WAN 28 (and possibly other networks). Although only a few ports are shown in Fig. 1, each line card typically has a plurality of ports. In the following description, port 36 is assumed to be an Ethernet port for the sake of brevity. As an alternative, some or all of the line cards may include tar type ports and may be connected to other types of networks, such as an Internet Protocol (IP) network. For example, in an alternate embodiment (not shown), the WAN 28 may include a Resilient Packet Ring (RPR) network, Interface. A feature of a network node that can be used to connect an Ethernet network to an RPR network is described, for example, in U.S. Patent Application No. 10 / 993,882, filed November 19, 2004. Additionally or alternatively, the line card 32 may connect a tunnel, such as a multi-protocol label-switching (MPLS) tunnel, over the WAN 28 via an appropriate label switching router within the WAN.

1, the specific ports 36 of the line cards 32 are connected by respective physical links to the switches 40 in the WAN 28 and these ports are coupled to the LAG group 38 do. Such a LAG group may serve a stance that is more than one VPLS. From the perspective of VPLS, a LAG group is a logical link with a total bandwidth (i.e. capacity) equal to the sum of the bandwidths of the respective physical links. At the physical level, for example, when a line card receives an incoming packet from the LAN 26 that should be sent to the WAN 28, the line card selects one of the ports in the LAG group for outgoing transmission of that packet . This port is typically chosen considering load balancing. For example, a line card may apply a hash function to a particular filter in the head of each incoming packet to select a port to transmit the packet to. The LAG group also provides built-in protection when one of the physical links in the group fails or becomes unavailable.

Figure 2 is a block diagram schematically illustrating details of a line card 32 within a node 30, in accordance with one embodiment of the present invention. The line card includes a plurality of ports 36, each port associated with a corresponding processing channel 50. (Although each channel 50 is shown as a functional block that is distinguished for simplicity, the actual channel may not be a distinct physical device, but rather a thread or process performed by a processing device serving a plurality of ports The port 36 of the first channel (channel 1), together with one or more ports on another line card (not shown in the figure), is connected to the LAG group 38, It is assumed to be connected to the switch 40 as part of the switch 40. [ The ports and associated channels connected to the LAN 26 or other nodes and media are similar in design and operation. The channel 50 includes an ingress path 54, and a packet processor 52, which includes an edge path 56. Packet processor 52 uses MAC FDB 58 for MAC learning and forwarding functions. The FDB is shared between the processing channels on the line card 32. This is formed and maintained according to the method described below with reference to Fig.

In the VPLS environment, each record in the FDB 58 corresponds to a particular MAC address belonging to a stance that is a particular VPLS. Optionally, a given VPLS stance may be partitioned into a number of virtual LANs (VLANs) that generally operate in a manner defined in the IEEE Standard 802.1Q described above. Thus, each record in the database is typically identified by a key including a MAC address, a VPLS identifier, and optionally a VLAN identifier or a VLAN group identifier (known as FID). When the header parameter of the incoming packet is recognized as matching the key, the corresponding record in the database indicates the output interface and other transmission parameters necessary for the node 30 to forward the packet to its destination. For a simple Ethernet interface, for example, a record can simply identify the port and line card to which the packet is to be transmitted. If a packet is to be forwarded through a LAG group, the record identifies the LAG group. The record also indicates that the contents of the record have been self-learned by the packet processor on that line card, or that the content has been received in a synchronization ("SYNC ") packet from another line card, Quot; flag. ≪ / RTI >

After receiving an incoming packet from the switch 40, the port 36 forwards the packet to the ingress path 54. The packet processor 52 identifies the VPLS (typically by a lookup and classification process based on any packet header field), and performs the in-comings (including the MAC destination address (DA), and optionally the VLAN identifier) Extracts the other key parameters from the packet, and uses the key to request the FDB 58. [ When a record is recognized, the packet processor adds to the packet a tag indicating the ingress port on which the packet was received, as well as the ingress port to which the packet should be forwarded. If the output interface indicated by the record is a LAG group, the packet processor selects one of the physical ports in the LAG group (e.g., using a hash function) and tags the packet for transmission on the selected port. The packet process then passes the tagged packet to the switching core 34, which forwards the packet to the egress path 56 of the appropriate port.

However, when the packet processor 52 receives a packet with no corresponding record in the database 58 for the key of the packet on the ingress path 54, the packet processor 52 tags the packet to flood it. In this case, the switching core 34 will pass the packet for transmission over all ports used by the stance, which is the VPLS (except for the ingress port where the packet was received). However, for each LAG group serving a stance that is a VPLS, the flooded packet is only sent through one port in the group.

The MAC database 58 and other forms of the learning process for nodes operating in an RPR environment, which may be applied to build databases, are described in the aforementioned U.S. Patent Application No. 10 / 993,882.

3 is a flow chart schematically illustrating a MAC learning method applied by a line card 32 in a node 30, in accordance with an embodiment of the present invention. The method is performed at step 60 by the packet processor 52 when processing packets in the egress path 56. Learning on this Grade is particularly advantageous for flooded packets because in this case multiple line cards can receive the packet and learn the interface association of the stanza that is the MAC source address (SA) and VPLS .

The packet processor 52 references the FDB 58 to retrieve the key parameters (MAC SA, VPLS stance, and optionally VLAN tag) of the packet on the greyscale path 56 in the key check step 62 do. If a record with this key does not already exist in the database, the packet processor creates a new record corresponding to that key, in an entry record step 64. [ The record indicates the interface on which the next packet received with the ingress path 54 along with this key should be forwarded, based on the input interface from which the current packet was received. If the packet for which the new record was created is a data packet, the packet processor sends the record to the SELF flag to indicate that it has learned the forwarding parameter from the forwarded packet via the greyscale path 56 of its own channel 50 Display. Otherwise, it indicates that the record is a SYNC record.

The packet processor then determines, at forwarding decision step 66, what the new record will do with the generated packet. If the packet is a data packet, the packet is forwarded to the appropriate output port in the forwarding step 68. Otherwise, the packet is simply discarded in the discard step 70.

("Sync") message to report each SELF entry that was generated in the FDB 58, at any regular interval (preferably less than the FDB aging time) To another line card (32). This message typically includes a message packet with the same header as the data packet forwarded by the node 30, with a special header field indicating that it is a synchronization message. The switching core 34 transmits this SYNC packet to the other line card in the same manner as forwarding the original data packet. However, the line card that received the packet will recognize the packet as a synchronization message and therefore will not process the packet internally without further forwarding it at step 64 (or at step 84, as described below) do.

At step 64, to process a SYNC packet with a new SA, each line card checks the stance, which is the VPLS identified in the packet. If the line card is not configured to serve a stance that is VPLS, simply discard the synchronization message. Otherwise, if there is no entry for the key field extracted from the SYNC packet, the line card adds the record to its FDB. In this case, as described above, the record has an indication that it is a SYNC entry, received from another line card.

1), the VPLS packet from the terminal 22 is forwarded by the node 30 to the terminal 24 via the switch 40 and the packet is forwarded to the LAG group 38, Lt; RTI ID = 0.0 > 36 < / RTI > However, all three line cards 32 with ports in the LAG group learn the port association of the MAC address of the terminal 22 by means of the SYNC packet sent by the line card to which the packet was forwarded. As a result, when the terminal 24 retransmits the packet to the terminal 22, all the line cards associated with the LAG group 38 can forward the packet to the appropriate interface for the terminal 22 without flooding. Learns the interface association of the MAC source address from other SYNC packets configured to support this VPLS stance (even if not in the same LAG group).

The advantage of using packets for distributing SYNC messages in the manner described above is that it allows the use of existing forwarding mechanisms within the node 30 without requiring additional control channels in the hardware. Alternatively, synchronization messages may be distributed among the line cards using a dedicated control channel. As an alternative or additionally, the line card may distribute each synchronization message only to the other line cards registered as serving the stanza that is the VPLS. However, the present inventor has found that the indiscriminate transmission of SYNC packets to all line cards simplifies the operation of the MAC learning mechanism, resulting in an additional additional communication burden. Additional savings can be achieved by sending multiple sync entries in a single packet. In this case, the above-described processing is simply repeated for a plurality of records in the same packet.

The aging mechanism is applied to the MAC database 58 to delete records that are no longer valid and to reserve space for new records. For this purpose, each record in the database has a time stamp that indicates when it was created or recently updated. A record with a given key is deleted from its database after a predetermined aging time without additional packets received with the same key, following that time stamp. Aging is applied to both SELF and SYNC records, and typically all have the same aging time. To prevent aging of the "live" record, the line card 32 refreshes the time stamp of the record in the manner described below.

3, when the packet processor 52 determines in step 62 that the FDB 58 already contains a record corresponding to the key of the current packet in its ingress path 56, In step 72, it is determined how to process the packet. If the packet is a data packet, the packet check step 74 checks the record in the FDB to determine if the current packet matches the record. That is, if there is no record in the FDB, the packet processor determines whether the current packet generated the same record as the existing record (that is, whether the existing record is a SELF record having the same port as the ingress port of the current packet) . If so, the packet process refreshes the time stamp of the record in the refreshing step 76 and forwards the packet to the appropriate output port in step 68.

On the other hand, if the packet processor determines in step 74 that the entry in the FDB 58 that matches the key of the current packet is a SYNC record, then the processor updates the record accordingly in an update step 78. As part of the update process, the packet process changes the SYNC indication in the record to SELF. Also, at step 78, after finding the key given by the packet in the output path 56, the packet processor 52 determines whether the input port of the packet is the interface currently recorded for that key in the database 58 May be found to be different. This kind of inconsistency may occur, for example, when the terminal 24 is moved to a different location, or when the network configuration changes due to an error or a new installation. In this case, the packet process records a new parameter in the SELF record overwriting the existing record.

The packet processor determines in an update decision step 80 whether changes made in the FDB record should be reported to another line card. If there are no changes to the interfaces listed in the record, the packet process simply forwards the data packet to the appropriate output port at step 68. [ However, if the interface has changed, the packet process sends a special SYNCUPDATE packet to the other line card at update step 82. This packet is similar to the above-described SYNC packet, but contains an additional "UPDATE" Typically, the SYNCUPDATE packet is sent immediately after updating the FDB record at step 76, without waiting for a predetermined time to transmit the SYNC packet. The data packet for which SYNCUPDATE is prompted is forwarded to the appropriate output port in step 68. [

Transmitting a specially marked SYNCUPDATE packet in this manner guarantees that every line card's MAC database is immediately updated as changes occur, avoiding race conditions between SYNC packets already transmitted between line cards with outdated information. The packet process that received a SYNCUPDATE packet with a different result than its own record data will change the record and change its entry state to SYNC, regardless of whether the record is a SYNC entry or a SELF entry, as described below .

Returning to step 72, if the packet processor 52 determines that the current packet is not a data packet (i.e., determines that the current packet is a SYNC or SYNCUPDATE packet) And checks to see if the existing record in the FDB 58 corresponding to the key of the current packet is a SYNC or SELF entry. In the case of a SYNC entry, the packet process updates the record in sync update step 86, if necessary. That is, if the interface indicated in the packet is different from the interface indicated in the existing record, the packet processor updates the record according to the packet. The packet process refreshes the timestamp of the record whether the record has changed or not. The packet is then discarded at step 70.

If the packet processor determines in step 84 that the existing record in the FDB 58 corresponding to the key of the current packet has been marked as a SELF record, then the processor checks the type of the packet in a type check step 88. If the current packet is SYNC, the packet process discards the packet, at step 70, because the SYNC packet does not overwrite the SELF entry. On the other hand, if the current packet is a SYNCUPDATE packet, the packet processor overwrites the SELF record in the FDB 58 in the SYNC update step 90 and marks the record as a SYNC entry. The packet is then discarded at step 70. [

In another embodiment (not shown) of the present invention, a redundant link between the node 30 and other network elements, such as a parallel link between the line card 32 and the switch 40, Can be used for protection in case one fails. This embodiment may also be advantageous for the above-described method for MAC database update and synchronization. More specifically, the standby line card can use a synchronized MAC database to generate a dummy data packet and transmit on each new active link, when activated to provide alternate service in the event of failure. After receiving such a packet, another device in the network learns to use the new active port. This mechanism of dummy packet transmission is described in detail in U.S. Patent Application No. 10 / 036,518, filed January 7, 2002, and published as US 2003/0208618 A1. To support this protection function, the FDB 58 is updated for the point-to-point service defined on the protected link, as well as updated for stances that are VPLS, as described above. In the latter case, the FDB record includes the MAC address and the connection ID, not the VPLS ID.

Although the above-described embodiments have been described with respect to particular exemplary network and equipment topologies and with reference to specific communication protocols, the principles of the present invention may be applied to various types and topologies of Layer 2 virtual private networks Can be similarly applied. It is to be understood, therefore, that the above-described embodiments are illustrative only and that the invention is not limited to the details shown and described herein. Rather, the scope of the present invention encompasses all possible variations and modifications to those skilled in the art having read the foregoing description, which is not disclosed in the prior art, as well as combinations and subcombinations of the various features described above.

Claims (24)

As a communication method, Configuring a network node having at least first and second line cards and a plurality of ports with respective first and second ports to operate as a distributed medium access control (MAC) bridge in a layer 2 data network; (LAG) group of parallel physical links between two endpoints in the layer 2 data network, the LAG groups being coupled together into a single logical link, the LAG group comprising a plurality of LAG ports and a plurality of joined member lines Having a card; Providing a respective forwarding database (FDB) for each of the member line cards to maintain a record associating a MAC address with a port of a plurality of ports of the network node; Receiving, from a MAC source address, an ingress port of the network node, a data packet specifying a MAC destination address on the layer 2 data network; Forwarding the received data packet at least at the network node to the first line card for transmission to the MAC destination address by forwarding the data packet to the MAC destination address via the first port ; Flooding the data packet through only one of the plurality of LAG ports when the MAC destination address is not present in the FDB; Checking the MAC source address of the data packet for a record in the FDB of the first line card; And If the FDB of the first line card does not include a record of the association of the MAC source address and the ingress port, generates a new record of the association and adds the new record to the FDB of the first line card And sending the association message to each member line card of the plurality of member line cards. 2. The method of claim 1, wherein transmitting the message comprises periodically sending a message at predetermined intervals to notify at least the second line card of a new association between the MAC address and the respective port Lt; / RTI > 2. The method of claim 1, wherein transmitting the message comprises receiving the message in the second line card, and in response to the message, if the record is not already present in the FDB of the second line card And adding the record of the association to the FDB of the second line card. 4. The method of claim 3, wherein adding the record further comprises: adding a first type of record added in response to a data packet transmitted on one port of the line card, and a response to a message received on another port of the line card Further comprising the step of displaying said record in each said FDB of each line card to distinguish the added second type of record. 5. The method of claim 4, wherein adding the record further comprises: Associating each of said records with an aging time of each; Refreshing the record in the FDB in response to an additional packet sent by the line card; And And if the record is not refreshed within the respective aging time, removing the record from each FDB. 2. The method of claim 1, wherein transmitting the message comprises transmitting a synchronization packet from the first line card to at least the second line card through the switching core of the network node. 7. The method of claim 6, wherein transmitting the sync packet comprises: if the record in the FDB associates the MAC source address with a port different from the one port on which the data packet was received, And sending a synchronization update packet to at least the second line card to indicate that the record has been changed. delete delete 8. The network node according to any one of claims 1 to 7, wherein the network node is configured to operate as a plurality of virtual MAC bridges in a layer 2 virtual private network (VPN) Wherein the record associating each port with the MAC address is maintained independently for each of the stans being the VPN. 11. The method of claim 10, wherein the VPN-in-stance is a stance that is one of a plurality of VPLSs served by the network node, and wherein the step of transmitting the message is to notify all the line cards serving the VPLS stance Identifying a stance that is the VPLS in the message. 8. The method according to any one of claims 1 to 7, wherein, following the step of transmitting the message, Forwarding additional data packets received from the additional MAC source address to the second line card for transmission on the network; Checking the additional MAC source address for the record in the FDB of the second line card; And Adding an additional record relating to the MAC source address to the FDB of the second line card in response to the additional data packet and transmitting an additional message to notify at least the first line card of the additional record Further comprising the steps of: As a node for network communication, Switching core; A plurality of ports; And (LAG) of physical links in parallel between two endpoints in the layer 2 data network, the nodes being coupled together into a single logical link for operating as a virtual medium access control (MAC) bridge in a layer 2 data network. And a plurality of member line cards for forwarding packets through the switching node with a plurality of LAG ports, Wherein the plurality of member line cards include at least first and second line cards, each line card including a respective port and for maintaining a record associating a MAC address with the respective port of the line card Each has a forwarding database (FDB) Wherein the line card receives a data packet specifying a MAC destination address from an MAC source address to an ingress line card, and wherein the ingress line card includes at least a first line for transmitting to the MAC destination address via the switching core, Said first line card checks said MAC source address of said data packet for a record in said FDB of said first line card, and said FDB of said first line card is said Adding the record to the FDB of the first line card and sending a message informing the at least the second line card of at least the second line card to the second line card if the record does not include an association record of the MAC address, And if the MAC destination address does not appear in the FDB, only one of the LAG ports Network communication node, characterized in that through the port, which is arranged to flood the data packet. 14. The system of claim 13, wherein at least the first line card is adapted to periodically transmit a message at predetermined intervals to notify at least the second line card of a new association between the MAC address and the respective port A node for network communication. 14. The method of claim 13, wherein, in response to the message, if the record is not already present in the FDB of the second line card, the second line card sends, to the MAC database of the second line card, To the network node. 16. The system of claim 15, wherein the record in each FDB of each line card is a record of a first type added in response to a data packet transmitted on one port of the line card, Is marked to distinguish the added second type of record in response to the received message. 17. The method of claim 16, wherein each aging time is associated with each said record, said line card refreshing said record in said FDB in response to an additional packet sent by said line card, And to delete the record in each FDB if it is not refreshed within the aging time. 14. The node of claim 13, wherein the message comprises a synchronization packet sent from the first line card to at least the second line card through the switching core. 19. The method of claim 18, wherein the line card associates the record in the FDB with the MAC source with a port different from the one port from which the data packet was received, if the record in the FDB of the first line card is changed And the synchronization packet includes a synchronization update packet indicating at least the second line card to indicate that the record has been changed. delete delete 20. A method according to any one of claims 13 to 19, wherein at least some of the line cards are configured such that the node is operating as a plurality of virtual MAC bridges in a Layer 2 virtual private network (VPN) Wherein the bridge is configured to serve a stance that is a respective VPN, and wherein the record associating the MAC address with the respective port is maintained independently for each stance that is the VPN. 23. The VPLS system of claim 22, wherein the stanza is a stance that is a VPLS of one of a plurality of VPLSs served by the network node, the stance being the VPLS notifies all of the line cards serving the VPLS stance Wherein the message is identified in the message. 20. The apparatus of any one of claims 13 to 19, further adapted to forward additional data packets received from an additional MAC source address to the first line card for transmission on the network, Checks the additional MAC source application for the record in the FDB of the second line card and, in response to the additional data packet, adds an additional record for the MAC source address to the FDB of the second line card And sends an additional message to notify the first line card of the additional record at least.

Images